kseltest: bpf: test_lwt_ip_encap.sh: Why -l used in test_gso

kseltest: bpf: test_lwt_ip_encap.sh: Why -l used in test_gso

[Prabhakar Kushwaha]

Hi All,

I am trying to understand test_gso test for IPv4 and IPv6 with following 
piece of code

       # listen on IPv*_DST, capture TCP into $TMPFILE
         if [ "${PROTO}" == "IPv4" ] ; then
                 IP_DST=${IPv4_DST}
                 ip netns exec ${NS3} bash -c \
                        "nc -4 -l -s ${IPv4_DST} -p 9000 > ${TMPFILE} &"
         elif [ "${PROTO}" == "IPv6" ] ; then
                 IP_DST=${IPv6_DST}
                 ip netns exec ${NS3} bash -c \
                        "nc -6 -l -s ${IPv6_DST} -p 9000 > ${TMPFILE} &"
                 RET=$?
         else
                 echo "    test_gso: unknown PROTO: ${PROTO}"
	fi

I have couple of queries around it

a) why -l is being used for reading listen on IPv*_DST with -s option.
I was looking at https://www.computerhope.com/unix/nc.htm, following has 
been mentioned:
-l :  Used to specify that nc should listen for an incoming connection 
rather than initiate a connection to a
remote host. It is an error to use this option in conjunction with the 
-p, -s, or -z options.
Additionally, any timeouts specified with the -w option are ignored.

b) Even if there is requirement of -l option to use. can we provide 
timeout option also. how? as -w dont work with -l.


I am facing an issue with Linux-5.3 Kselftest where even if 
test_lwt_ip_encap.sh "exit", bpf/runner.sh is not
running next test case. It just wait until CRTL + c is pressed.
If I comment above code things work fine.


Please sugggest

--prabhakar (pk)

Re: kseltest: bpf: test_lwt_ip_encap.sh: Why -l used in test_gso

[Liu Yiding]

Hi, Kushwaha.


On 10/13/19 9:59 AM, Prabhakar Kushwaha wrote:
> Hi All,
>
> I am trying to understand test_gso test for IPv4 and IPv6 with 
> following piece of code
>
>       # listen on IPv*_DST, capture TCP into $TMPFILE
>         if [ "${PROTO}" == "IPv4" ] ; then
>                 IP_DST=${IPv4_DST}
>                 ip netns exec ${NS3} bash -c \
>                        "nc -4 -l -s ${IPv4_DST} -p 9000 > ${TMPFILE} &"
>         elif [ "${PROTO}" == "IPv6" ] ; then
>                 IP_DST=${IPv6_DST}
>                 ip netns exec ${NS3} bash -c \
>                        "nc -6 -l -s ${IPv6_DST} -p 9000 > ${TMPFILE} &"
>                 RET=$?
>         else
>                 echo "    test_gso: unknown PROTO: ${PROTO}"
>     fi
>
> I have couple of queries around it
>
> a) why -l is being used for reading listen on IPv*_DST with -s option.
> I was looking at https://www.computerhope.com/unix/nc.htm, following 
> has been mentioned:
> -l :  Used to specify that nc should listen for an incoming connection 
> rather than initiate a connection to a
> remote host. It is an error to use this option in conjunction with the 
> -p, -s, or -z options.
> Additionally, any timeouts specified with the -w option are ignored.
>
> b) Even if there is requirement of -l option to use. can we provide 
> timeout option also. how? as -w dont work with -l.
>
>
> I am facing an issue with Linux-5.3 Kselftest where even if 
> test_lwt_ip_encap.sh "exit", bpf/runner.sh is not
> running next test case. It just wait until CRTL + c is pressed.


I meet same issue. "make run_tests -C bpf" bpf/runner.sh hang on 
test_lwt_ip_encap.sh. And when i skip it,

test_tc_tunnel.sh also has this problem.


> If I comment above code things work fine.
>
>
> Please sugggest
>
> --prabhakar (pk)
>
>
-- 
Best regards.
Liu Yiding

Re: kseltest: bpf: test_lwt_ip_encap.sh: Why -l used in test_gso

[Prabhakar Kushwaha]

On 10/14/2019 12:04 PM, Liu Yiding wrote:
> Hi, Kushwaha.
> 
> 
> On 10/13/19 9:59 AM, Prabhakar Kushwaha wrote:
>> Hi All,
>>
>> I am trying to understand test_gso test for IPv4 and IPv6 with 
>> following piece of code
>>
>>       # listen on IPv*_DST, capture TCP into $TMPFILE
>>         if [ "${PROTO}" == "IPv4" ] ; then
>>                 IP_DST=${IPv4_DST}
>>                 ip netns exec ${NS3} bash -c \
>>                        "nc -4 -l -s ${IPv4_DST} -p 9000 > ${TMPFILE} &"
>>         elif [ "${PROTO}" == "IPv6" ] ; then
>>                 IP_DST=${IPv6_DST}
>>                 ip netns exec ${NS3} bash -c \
>>                        "nc -6 -l -s ${IPv6_DST} -p 9000 > ${TMPFILE} &"
>>                 RET=$?
>>         else
>>                 echo "    test_gso: unknown PROTO: ${PROTO}"
>>     fi
>>
>> I have couple of queries around it
>>
>> a) why -l is being used for reading listen on IPv*_DST with -s option.
>> I was looking at https://www.computerhope.com/unix/nc.htm, following 
>> has been mentioned:
>> -l :  Used to specify that nc should listen for an incoming connection 
>> rather than initiate a connection to a
>> remote host. It is an error to use this option in conjunction with the 
>> -p, -s, or -z options.
>> Additionally, any timeouts specified with the -w option are ignored.
>>
>> b) Even if there is requirement of -l option to use. can we provide 
>> timeout option also. how? as -w dont work with -l.
>>
>>
>> I am facing an issue with Linux-5.3 Kselftest where even if 
>> test_lwt_ip_encap.sh "exit", bpf/runner.sh is not
>> running next test case. It just wait until CRTL + c is pressed.
> 
> 
> I meet same issue. "make run_tests -C bpf" bpf/runner.sh hang on 
> test_lwt_ip_encap.sh. And when i skip it,
> 
> test_tc_tunnel.sh also has this problem.
> 
> 

does this means, the problem is independent of "nc" and has no relation 
with nc usage with -l option.

has anyone find it's root-cause..

-prabhakar (pk)

Re: kseltest: bpf: test_lwt_ip_encap.sh: Why -l used in test_gso

[Liu Yiding]

On 10/14/19 10:29 PM, Prabhakar Kushwaha wrote:
> On 10/14/2019 12:04 PM, Liu Yiding wrote:
>> Hi, Kushwaha.
>>
>>
>> On 10/13/19 9:59 AM, Prabhakar Kushwaha wrote:
>>> Hi All,
>>>
>>> I am trying to understand test_gso test for IPv4 and IPv6 with 
>>> following piece of code
>>>
>>>       # listen on IPv*_DST, capture TCP into $TMPFILE
>>>         if [ "${PROTO}" == "IPv4" ] ; then
>>>                 IP_DST=${IPv4_DST}
>>>                 ip netns exec ${NS3} bash -c \
>>>                        "nc -4 -l -s ${IPv4_DST} -p 9000 > ${TMPFILE} &"
>>>         elif [ "${PROTO}" == "IPv6" ] ; then
>>>                 IP_DST=${IPv6_DST}
>>>                 ip netns exec ${NS3} bash -c \
>>>                        "nc -6 -l -s ${IPv6_DST} -p 9000 > ${TMPFILE} &"
>>>                 RET=$?
>>>         else
>>>                 echo "    test_gso: unknown PROTO: ${PROTO}"
>>>     fi
>>>
>>> I have couple of queries around it
>>>
>>> a) why -l is being used for reading listen on IPv*_DST with -s option.
>>> I was looking at https://www.computerhope.com/unix/nc.htm, following 
>>> has been mentioned:
>>> -l :  Used to specify that nc should listen for an incoming 
>>> connection rather than initiate a connection to a
>>> remote host. It is an error to use this option in conjunction with 
>>> the -p, -s, or -z options.
>>> Additionally, any timeouts specified with the -w option are ignored.
>>>
>>> b) Even if there is requirement of -l option to use. can we provide 
>>> timeout option also. how? as -w dont work with -l.
>>>
>>>
>>> I am facing an issue with Linux-5.3 Kselftest where even if 
>>> test_lwt_ip_encap.sh "exit", bpf/runner.sh is not
>>> running next test case. It just wait until CRTL + c is pressed.
>>
>>
>> I meet same issue. "make run_tests -C bpf" bpf/runner.sh hang on 
>> test_lwt_ip_encap.sh. And when i skip it,
>>
>> test_tc_tunnel.sh also has this problem.
>>
>>
>
> does this means, the problem is independent of "nc" and has no 
> relation with nc usage with -l option.
>
> has anyone find it's root-cause..


I think the root cause is related to kselftest/prefix.pl, you could try 
to kill process prefix.pl, then runner.sh will run next test.


>
> -prabhakar (pk)
>
>
>
>
-- 
Best regards.
Liu Yiding